808
Macromolecules 1980,13, 808-815
a close simulation of the spectrum in dimethylformamide as well, with appropriate detailed changes in absolute chemical shifts and line widths. It was necessary to assign variable line widths to represent the differing degrees of line broadening from unresolved heptad stereosequences. The given stereosequence assignments are consistent with a bias toward racemic placement, which increased with decreasing temperature (Figure 1). However, m and r could just as easily be interchanged for an adequate simulation with a P(m) value of 0.58. Definitive assignments cannot be obtained without a study of the appropriate model compounds. The areas predicted by the present simulation deviate most noticeably for the rmmm and mrrm peaks. A better simulation is possible by assuming Markov 1statistics with P(mlm) = 0.526 and P(rlm) = 0.353. The quantitative significance of the statistical treatment must remain in doubt, however, owing to the tentative nature of the assignments. An interleaved Bernoulli stereosequence propagation model6would perhaps be the most appropriate for the present system, where the probability of one dyad is related to the isomerization of the terminal radical and
the probability of the next dyad is related to the addition of a monomer unit. The absence of crystallinity in PFM3 is hardly surprising given its degree of stereoirregularity. 19FNMR is an excellent probe of the microstructure, though a detailed understanding of the ordering of the stereosequences according to chemical shift, and the dependence of the ordering and spacing on solvent, remains a challenging problem. Acknowledgment. The author is grateful to Dr. F. A. Bovey for his interest in this work. References and Notes (1) R. C. Ferguson and E. G. Brame, Jr., J . Phys. Chem., 83, 1397-1401 (1979),and references therein. (2) T. Yagi, Polym. J., 11, 353-8 (19791,and references therein. (3) W.S. Durrell, G. Westmoreland, and M. G. Moshonas, J. Polym. Sci., Part A , 3, 2975-82 (1965). (4) H.Yane, Ph.D. Dissertation. Dartmouth Colleee. NH. 1976. (5) M. K. Lkkapeddi and H. K. Reimschuessel, M&rbmoZecules, 11, 1067-74 (1978). (6) H. L.Frisch, C. L. Mallows, and F. A. Bovey, J . Chem. Phys.,
45, 1565-77 (1966).
Photooxidation of Di-n-butyl Sulfide Using Sensitizers Immobilized in Polymer Films Richard A. Kenley,* Norman A. Kirshen, a n d Theodore Mill SRI International, Menlo Park, California 94025. Received December 12, 1979
ABSTRACT The photooxidation of di-n-butyl sulfide (Bu,S) to di-n-butyl sulfoxide (Bu2SO)was investigated in methanol solution by using various dyes as singlet-oxygen sensitizers. The rates of BuzSO formation are zero order in [Bu2S]and in the concentration of dissolved oxygen and thus serve as a measure of sensitizer efficiency. The quantum yields for BuzSO formation, using sensitizers dissolved in methanol, were 1.17,1.31, 0.201,0.024,0.0081,0.0064, and C0.0031 for rose bengal (RB), tris(2,2’-bipyridyl)ruthenium(II)[Ru(bpy),], and malachite green, saffranine 0 (SO), rhodamine B, victoria blue B, l-amino-4-hydroxyanthraquinone, respectively. RB, SO, and Ru(bpy), were immobilized in poly(viny1 chloride) [PVC] films and RB was immobilized in cellulose acetate (CA) films. Bu2S photooxidation quantum yields for the films were 1.2-1.8, 0.019-0.024,0.026-0.044, and 0.018-0.020 for RB/CA, RB/PVC, Ru(bpy),/PVC, and SO/PVC, respectively. Photodegradation of immobilized sensitizers to colorless products was also monitored. RB/CA, RB/PVC, SO/PVC, and Ru(bpy),/PVC films were 50% photodegraded after irradiation for 0.15,5, 7, and 320 h, respectively. The results indicate that polymer films retard transport of solvent, dissolved oxygen, and/or substrate to sites occupied by the photosensitizer and that there is an attendant decrease in the photooxidation quantum yield relative to that of photosensitizers in solution. Substrate photooxidation and sensitizer photodegradation are parallel processes, probably involving energy transfer from sensitizer to 302 in the former case and electron transfer to form sensitizer radical cation and superoxide in the latter.
Molecular oxygen in its first (singlet) electronic excited undergoes a variety of reactions, including state (lo2) 1,4-cycloaddition to dienes, the “ene” reaction with alkyl-substituted olefins to produce hydroperoxides, 1,2addition to electron-rich olefins, oxidation of phenols, and oxidation of dialkyl sulfides to dialkyl su1foxides.l These reactions are of interest both because they have useful synthetic applications and because lo2has been implicated as the reactive species in photooxidations of biological systems2 and of synthetic high polymer^.^ Recently, attention has focused on photosensitized ‘02 reactions employing sensitizers immobilized on various solid s ~ p p o r t s . ~ -Immobilized l~ sensitizers offer an advantage over dissolved photosensitizers because they can be employed in solvent systems in which the sensitizer is normally insoluble. Also, immobilization of a sensitizer limits side reactions between sensitizer and oxidizable substrate and facilitates the separation of the photosensitizer from reaction products.
In general, three approaches have been taken in the design of immobilized photosensitizers. In ~ n e , the ~ ~ J ~ sensitizer is covalently bound to cross-linked polystyrene beads, via reaction of the sensitizer with chloromethylated polystyrene. A primary advantage of this approach is that the sensitizer is highly resistant to separation from the support material, thus permitting use of the photosensitizer in a wide variety of solvent systems. Although the scope of this approach is unlimited in principle, relatively few photosensitizers can be linked to polymer supports simply and conveniently. Also, absorption and scattering of incident radiation by the polymer beads complicates the use of covalently bound photosensitizers in studies (such as determination of photooxidation quantum yields) in which the intensity of the absorbed radiation must be well-known. A second approach is immobilization of a photosensitizer by absorption onto a solid support such as silica gel.8~9216 This method is convenient, but the sensitizer can be de-
0024-9297/80/2213-0808$01.00/00 1980 American Chemical Society
Photooxidation of Di-n-butyl Sulfide 809
Vol. 13, No. 4, July-August 1980
Table I Production of Bu,SO and Disappearance of Bu,S from Photooxidation of 4.7 x lo-' M Bu,S in Methanol with Ru(bpy), Sensitizef time, s
[Bu,S],, M
10, x
103 x [Bu,SO],, M
Bu,S convrsn,b %
0 60 120 180 240 360
4.70 4.41 4.09 3.68 3.39 2.86
0.1 3.6 6.7 10.1 12.8 18.2
0.2 7.7 14.2 21.4 27.2 38.7
Sensitizer concentrations before and after reaction were 8.3 Average values.
X
l o s (+d[Bu,SO] / dt), M s-'
lo5(-d[Bu,S] /
6.00 5.58 5.61 5.33 5.06 5.52 * 0.35c
4.83 5.08 5.67 5.46 5.11 5.23 t 0.33c
dt), M s-*
M, identical within experimental uncertainty ( * 3%).
% conversion = 100 X [Bu,SO]~/[BU,S]o.
tively, and Bu2S+OO-is the persulfoxide, believed to be sorbed from the support by polar solvents. Light scattering an intermediate in the reaction. by the support material is also an inherent difficulty. A third approach is incorporation of a photosensitizer Scheme I into a polymer matrix by dissolving the photosensitizer and ki polymer in a suitable solvent, casting a thin film of the 'S + hv e 'S* mixture on a glass plate, and allowing the solvent to k-1 e ~ a p o r a t e . ' ~ * ' ~The - ' ~ number of photosensitizers and polymers that can be employed is large. Moreover, cast films are highly transparent and thus can be used to prepare sensitizer films whose spectral properties are well k3 3S* 302 IS IO2 characterized. This technique has been used to probe the diffusivity and reactivity of '02in polymer m a t r i ~ e s ' ~ J ~ J ~ J ~ k4 and to investigate the relative efficiency and susceptibility 3s*-+ 1s to photodegradation of various photosensitizer dyes.13J5 k6 Zweig and Hender~on'~ immobilized a series of 26 comB u ~ S+ '02 Bu~S'OOmercially available dyes in cellulose acetate (CA) films and investigated the relative abilities of the dyes to sensitize k0 the photooxidation of a known acceptor (dimethylB u ~ S+ Bu~S+OO- 2BuzSO anthracene) also incorporated into the films. These aukl thors characterized several of the dyes as efficient sensiBuzS + Bu2S+OO- 2Bu2S + 302 tizers and resistant to photodegradation. k0 In our laboratory, we have studied several parameters Bu~S+OO- B u ~ S O ~ that govern the ability of sensitizers immobilized in polymer films to effect photooxidation of organic subk9 strates in solution. We chose methanol as a representative IO2 302 solvent phase and di-n-butyl sulfide (Bu2S) as a wellBu2S+OO-is partitioned among reaction with Bu2S to chara~terized'~-'~ '02 acceptor. All experiments used yield di-n-butyl sulfoxide (Bu2SO)(reaction 61, quenching monochromatic light of known intensity, thus permitting with Bu2S (reaction 7) to revert to starting material plus determination of photooxidation quantum yields. Initially 302, and reaction 8, yielding di-n-butyl sulfone (Bu2S02). we examined the quantum yields for oxidation of Bu2S by Similarly, the excited sensitizer states are partitioned using various sensitizers dissolved in methanol, employing between decay to ground state (reactions -1 and 4) and six dyes studied by Zweig and Henderson and, for comeither intersystem crossing (reaction 2) or energy transfer parison, tris (2,2'-bipyridyl) ruthenium (11) fluoroborate to 302 (reaction 3). The '02also reacts by parallel path[ R ~ ( b p y ) ~ ] .We ~ *then ~ ~ immobilized the three most efways, viz., oxidation of Bu2Sto Bu,S+OO- (reaction 5) and ficient sensitizers in thin films of various polymers for decay to 302 (reaction 9). The overall kinetic behavior of further study. Finally, sensitizers immobilized in polythe photosensitized oxidation will, of course, depend on (vinyl chloride) [PVC] and CA films were investigated with the relative rates of reactions 1-9. By applying the respect to photooxidation quantum yields and rates of steady-state approximation to 13S] and reactions 1-4 we sensitizer photodegradation. We examined several paderive eq 10,where 4% is the quantum yield for production rameters of the sensitizer-polymer system that will govern the usefulness of these materials as practical photooxidation sensitizers, including sensitizer efficiency and quantum yield, polymer composition, and rates of transport of oxygen and solvent into the film. of singlet oxygen, E is the absorbed photon concentration in einstein L-I, and I, (=dE/dt) is the intensity of light Results a n d Discussion absorbed by 'S. When k 3 [ 3 0 2 ]> k4,23eq 10 reduces to Kinetics of Photooxidation of Bu2S. The photooxidation of BuzS is a well-behaved and well-characterized ~ystem.'~-'~ Bu2S reacts rapidly with singlet oxygen ('0,) to form first the persulfoxide, and, on reaction with a second molecule of Bu2S, two molecules of Bu2S0. Other In this case, 402 is independent of the dissolved-oxygen processes can compete, and the detailed scheme is shown concentration and is a measure of k , / ( k - , + k 2 ) ,i.e., the in Scheme I, where IS, ' S * , 3S*are the sensitizer singlet efficiency of intersystem crossing for the sensitizer in ground, singlet excited, and triplet excited states, respecquestion. Application of the steady-state approximation
+
-
+
-
-
+
-
+
810 Kenley, Kirshen, and Mill
Macromolecules
to the kinetics of Bu2S0 formation gives eq 12, where +d[Bu&O] /dt
is the quantum yield for BuzSO production and 40, is given by eq 10. The factor of 2 in eq 12 is dictated by the 2:l stoichiometry of reactions 5 and 6. For diethyl sulfide photooxidation in methanol, Foote and Peters18 demonstrated that k5 >> k,, and that (k6 + k7) >> k8. Therefore, providing that [Bus] remains sufficiently high, eq 12 reduces to 4Bu2S0 =
24OZ[k6/(k6 + k7)I
(13)
Under these conditions, the rate of Bu2S0 formation is zero order in both [,02] and [Bu2S],and Bu2S0is the only important product. Also, 4Bu2S0 is directly proportional to 4 0 so ~ 4 ~ can~be used ~ to ~ calculate 0 the relative efficiencies of various sensitizers. Experimental Results To ensure that our photooxidation system was wellbehaved, we first studied the kinetics of the photooxidation of 4.7 X M BuzS, using 8.3 X M Ru(bpy), dissolved in methanol as the sensitizer and a medium-pressure Hg lamp filtered to pass the 546- and 578-nm emissions. The production of Bu2S0 and the disappearance of Bu2S were both determined by high-pressure liquid chromatography. The results are shown in Table I (actinometry was not performed in this experiment). The rates for Bu2S disappearance and Bu2S0 formation are identical, within experimental uncertainty. This demonstrates that BuzSO is the major product and that Bu#O2 formation is neglible. Also, Table I shows that the reaction rates are invariant for conversions of Bu2S0 as high as 39%, indicating that the reaction is zero order in Bu2S. To demonstrate the independence of the reaction on [,02], the photooxidation of 4 X M Bu2S was run in methanol saturated with 1 atm of air and with 1 and 2 atm of oxygen. Rose bengal (RB) was the sensitizer employed, and actinometry was performed so that eq 14 and 15 could be used to calculate quantum yields for Bu2S0 formation +d[BuZSO]/dt 4BuzSO
=
-
dE/dt
+d[Bu,SO]/dt Za
(14)
is the quantum yield for production of BU~SO, dEfdt refers to the rate of absorption of light (in einstein L-I s-l), and I,, the absorbed intensity, is calculated from
I, = z0546(1- IO-^^)
+1
~ ~- 10-~678) ~ ~ ( 1 (15)
Here, I , is the total absorbed intensity, and Z0578 are the incident light intensities, and A546and A578 are the absorbances of the photosensitizer at 546 and 578 nm, respectively. Data for this series of experiments are summarized in Table 11. From the table, it is evident that the values of 4BuB0 for 1and 2 atm of oxygen are identical (the average value is 4Bu so = 1.08 f 0.10), but the value of ~ B , , ~ S Ofor 0.2 atm o f 0 2 is only 0.62. This suggests that the Bu2S photooxidation is zero order in [,02]if the solution is saturated with at least l atm of oxygen. This was the case in all subsequent experiments. We also measured the rates of oxygen consumption and of Bu2S disappearance for the RB-sensitized photooxidation in methanol and observed a Bu2S:02stoichiometry of 2.08 f 0.10, as required by reactions 5 and 6.
Table I1 Photooxidation of Bu,S in Air- or Oxygen-Saturated Methanol with Rose Bengal Sensitize+ 10' x 105 x 10 la? b [Bu,S],, (+d[Bu,SO]/ einstein conditions M dt), M s-' L-' s-' @Bu.SOc air (1 atm) 3.95 4.42 7.06d 0.626 oxygen 3.58 8.33 7.06d 1.18 ( 1 atm) oxygen 4.53 11.8 ll.Oe 1.07 ( 1 atm) ll.Oe 0.973 oxygen 4.53 10.7 ( 2 atm)
RB concentrations before and after reaction were 7.7 M, identical within experimental uncertainty (* 3%). I, calculated from eq 15. B~,SO calculated I,s47= 8.52 x l ( T s einstein L-'s-'. e I,547 from eq 14. = 13.3 X einstein L-' s-'. a
X
Table I11 Pertinent Spectral Characteristicsu of Selected Sensitizers
rose bengal (acid red 9 4 ) saffranine orange (basic red 2) 1-amino-4-hydroxyanthraquinone (disperse red 1 5 ) tris(2,2'-bipyridyl)ruthenium(II) fluoro borate malachite green (basic green 4 ) victoria blue B (basic blue 26) rhodamine B, (basic violet 10)
555
82000
6 5 0 0 0 11 000
528
53 900
28 600
1 260
522
8 230
7 350
4 420
448
7 000
360
20
620 165 000
20 000 48 000
604
28 200 56 000
64 000
554 1 0 9 1 0 0 1 0 9 0 0 0
9560
(I Determined experimentally from UV-visible absorption curves for solutions of accurately known concentrations of sensitizer in CH,OH.
In summary, the rates and products of the photosensitized Bu2S oxidation in methanol are in accordance with the mechanism outlined above. Therefore, rates of Bu2S photooxidation determined under conditions in which I , is known can be used as valid measures of relative sensitizer efficiency, as defined by eq 11 and 13. Efficiencies of Photosensitizers in Methanol Solution. Since we were interested in examining the effect of sensitizer efficiencies in lozoxidations, we chose six organic dyes studied by Zweig and Henderson15 and reexamined their behavior under conditions in which the intensity of absorbed radiation could be determined in a straightforward manner. We also investigated the behavior of Ru(bpy), as an example of an inorganic photosensitizer. The sensitizers studied and pertinent spectral data are listed in Table 111. Photooxidations with the sensitizers dissolved in methanol were performed by using 5 X M BuzS as the substrate and 546- and 578-nm emissions of the Hg lamp as the source of illumination. Sensitizer concentrations were adjusted to give an absorbance of 0.50 at 546 nm, the absorbances at 578 nm were calculated from the data in Table 111, and absorbed intensities were calculated from eq 15. In each experiment we determined the amount of Bu2S0formed after 120 s of irradiation, and values of ~B,,,sowere calculated from eq 14. The data, summarized in Table IV, reveal that the sensitizer efficiencies vary by a factor of >400. This result is in marked contrast to the findings of Zweig and Hen-
Vol. 13, NO.4, July--August 1980
Photooxidation of Di-n-butyl Sulfide 811
Table IV Rates (d[Bu,SO]/dt) and Quantum Yields ( @ B S~ O )of Bu,SO Production $om Photooxidations of 5.0 X l o - * M Bu,S in Methanol with Various Sensitizers 105 x 1051,; 105 x [sensieinstein (d[Bu,SO]/ sensi ti z e P t i ~ e r ]M , ~ L-' s - ' dt),d M s" 0.77 7.30 8.58 rose bengal (RB) 4.64 6.08 tris( 2,2'-bipyridyl)ruthenium(II)fluoroborate [Ru(bpy),] 1.75 6.37 1.28 saffranine 0 (SO) 0.46 6.71 0.163 rhodamine B 1.77 12.5 0.101 victoria blue B 6.8 9.64 0.0620 1-amino-4-hydroxyanthraquinone malachite green 2.5 12.8 < 0. 04f
@Bu,SOe
1.17 1.31 0.201 0.024 0.0081 0.0064 < 0.0031
In all cases sensitizer concentrations before and after irradiation See Table I11 for sensitizer spectral characteristics. were identical, within experimental uncertainty (i 3%);concentrations chosen to ' la calculated from eq 5 4 6 = 0.50. einstein L" s-l and Io5'* = 7.29 x 10+ einstein L-'s-'. 2Vde[&u,SO]/dt= [ B u , S 0 ] / 1 2 0 s , for 12015;1,546= 8.78 x M s irradiation. e @ B u , S O calculated from eq 14. f Bu,SO formation not observed within limits of detectability ( 5 X Bu,SO). Table V derson,15who observed all six organic dyes to be of comPhysical and Spectral Characteristics of Sensitizer Films parable efficiency as photosensitizers. It thus appears that the observed efficiency of immobilized sensitizers is not Ru(bpy)/ SO/ RB/ RB/ PVC PVC PVC CA necessarily an accurate measure of inherent efficiency of the sensitizers. As described below, the limiting effect of film thickness, 2.5 2.5 2.5 2.5 polymer on mass (i.e., oxygen) transport can affect apc m x lo', 0.085 0.24 0.26 0.20 absorbance of film parent efficiencies in this manner. a t 546 nm (A,) Our values of 4Bu2S0 = 1.2 for RB and $Bu2S0 = 1.3 for sensitizer concn,a 94.0 3.3 1.6 1.2 Ru(bpy),, can be compared to literature values of 240?= M X 10-3 1.6 for RBlctZ4and 240 = 1.7 for Ru(bpy),, for generation fraction of 546-nm 0.18 0.42 0.45 0.37 of lo2in methanol. $he relative values of 4Buso(RB)/ light absorbedb ~ B ~ ~ s o ( R u ( ~=P0.90 Y)~ ) ~ ~ o , ( R B ) / ~ ~ o , ( R ~= ~ P Y ) B ) and is the sensitizer A , , , / €5 4 6 (2.5 X l o - ) cm), where 0.94 compare favorably, although our absolute values are extinction coefficient at 546 nm (taken from Table 111). approximately 25% lower than the values predicted from bFraction absorbed = l - l O - A . the efficiency of sensitizer energy transfer to ,02. Cassagrande et al.19 also noted that, for the photooxidation ished as the polymer film formed. This indicates acidic of BuzS in ethanol sensitized by methylene blue, the value impurities in the poly(buty1 methacrylate) that become of 4Bu&o= 0.61 was significantly lower than the reportedz5 more concentrated as the solvent for the polymer evapovalue of 240, = 1.04 for methylene blue. These observarates. When RB/poly(butyl methacrylate) films were imtions are explained by reactions 6 and 7 and eq 13. Acmersed in methanol, they became cloudy and shriveled. cording to the proposed mechanism for alkyl sulfide oxiSensitizer films made from cellulose acetate (CA) were dation, the intermediate persulfoxide can react with alkyl very smooth and highly transparent. When RB/CA films sulfide to produce 2 mol of sulfoxide, or the persulfoxide were immersed in methanol, no leaching of dye into the can be quenched by sulfide to starting materials and solvent occurred over several hours. Ru(bpy), and SO, ground-state molecular oxygen. As indicated in eq 13, however, were completely leached out of CA in minutes. therefore, the efficiency of Bu2S0oxidation will differ from PVC (actually a vinyl chloride-vinyl acetate copolymer; the maximum value of 240 by the factor k 6 / ( k , + k 7 ) . see experimental details) was the only material investigated From our data with RB and Ru(bpy), as sensitizers, we that could be used to make satisfactory films containing calculate k,/(k6 + h7) = 0.76 f 0.01 and k6/k7 = 3.2 for all three sensitizers. Leaching of the dyes from PVC films BuzS in methanol. From the results of Cassagrande et al., was negligible except with very long reaction times. Dry these values for BuzS in ethanol are k 6 / ( k 6 + k7) = 0.59 sensitizer/PVC films were stippled, resembling an orange and k 6 / k 7 = 1.44. Both calculations indicate that the peel. Some blushing (cloudiness) of the films occurred on oxidation reaction predominates over the quenching reprolonged irradiation, but this did not affect the efficienaction, but the differences in the ratio k 6 / k 7 are rather cies of the photosensitizer (see Table VI). large. The reasons for this difference are unclear, although On the basis of these limitations, we investigated the the quenching:oxidation ratio is known to be solvent-derates of photodegradation of the sensitizers and rates of pendent.18 photooxidation of BuzS, using RB/CA, RB/PVC, RuPhotooxidation of Bu2SUsing Sensitizer Films. We (bpy),/PVC, and SO/PVC films. Table V lists pertinent chose the three most efficient sensitizers from Table data for the four sensitizer films employed. IV-RB, Ru(bpy),, and saffranine 0 (SO)-as candidates The sensitizer photodegradation and BuzS photofor further study. We immobilized the sensitizers in thin oxidation experiments were performed concurrently. Each films made from various polymers (see Experimental film was cut into strips, and each strip was immersed in Section for details). 2 mL of a methanol solution of 5 X M BuzS in a Polystyrene films were transparent when undyed but photolysis tube. The tubes were then irradiated for timed became cloudy when RB was incorporated into the film. intervals with a Hg lamp filtered to pass only the 546-nm When RB/polystyrene films were immersed in methanol, emission. The film was removed after irradiation and the absorbance of the film was determined. At the same time, the sensitizer did not leach into the solvent. Ru(bpy),, however, was leached out of polystyrene films within a few the concentration of Bu2S remaining in the tube was minutes. When RB was incorporated into a poly(buty1 quantitated and the rate of BuzS photooxidation calcumethacrylate) film, the color intensity of the dye diminlated. Actinometry was performed for these experiments
812 Kenley, Kirshen, and Mill
Macromolecules
Table VI Photooxidation of Bu,S in Methanol Using Various Photosensitizer Filmsa sensitizer/filmb time, h @Bu,SOC
48 0 3 6d
SOIPVC RBIPVC ~.
0.034 0.018 0.020
0
~
2
4 8 16d 0 0.25 0.5 1.0
RB/CA
0.019 0.024 0.024 0.057 1.2 1.8 1.4
Experiments performed simultaneously with measurement of rates of photodegradation of sensitizer films. See Table V for details concerning films. @ B ~ 2 , ([ABu2S]/AtY,-*;see text. Photooxidations could not be followed to longer times due to leaching of sensitizer from film into methanol.
0.6
I
0 0
10
20 TIME
1 30
- hour
40
50
Figure 1. Fraction of sensitizer remaining ( A , / A o )vs. time for photolysis of Ru(bpy)/PVC (O),SO/PVC (@), RB/PVC (w), and RB/CA (0)films in methanol containing approximately 5 x lo-' M Bu2S. Values in parentheses are the times required for A,/Ao to reach 0.50. and values of $Buzsowere derived. The results of the photodegradation experiments are depicted in Figure 1. The ratio of A, (absorbance of the film after irradiation for time t ) to A. (initial absorbance of the film) is a measure of the fraction of sensitizer remaining. It is evident that the four immobilized sensitizers have greatly different susceptibilities to photodegradation. For example, the Ru(bpy),/PVC film is about 60 times more resistant to photodegradation than the RB/PVC film. The figure also shows that photodegradation rates for a given sensitizer depend on the polymer used for immobilization. The RB/PVC film, for example, is 50% degraded in 5 h, whereas the RB/CA material undergoes a similar degree of fading in about 10 min, a 30-fold difference in bleaching rate. Analogous behavior was noted for the rates of Bu2S oxidation, summarized in Table VI. The rates of Bu2S oxidation, A[Bu2S]/At were calculated for each time point. Values of $ B u r n were calculated from the expression &a ,0 = (A[Bu,S]/At)Z['. This calculation requires the rate of
~
Bu2S consumption to equal the rate of Bu2S0 formation, a valid approximation according to the data in Table I. Values of Z, at 546 nm were taken from Figure 1. Since RB and SO were noticeably degrading during the experiments, the I, values continually decreased. Thus, the calculation of 4Buzsoabove is not rigorously correct, and the quantum yield in each time interval is actually somewhat higher. However, the values of $BuzSO so obtained provide good approximations that are useful in evaluating the performance of the sensitizers. As Table VI shows, all three sensitizers immobilized in PVC films exhibited $B,,~so values of 0.02 to 0.06. Quantum yields for Bu2S0formation sensitized by R ~ ( b p y )RB, ~ , and SO in solution were 1.3, 1.2, and 0.2, respectively. In contrast, RB/CA films afforded $Bu2S values of 1.2-1.8, approximately equal to the quantum yields obtained with RB in solution. Since RB/CA films also faded about 30 times faster than did RB/PVC films, it is evident that incorporation of the sensitizers into different films affects photodegradation and photooxidation in parallel. The best explanation for these observations seems to be ~ that methanol diffuses through the polymer films, carrying along both oxygen and Bu2S. During passage through the films, 302 is converted to lo2and reacts with Bu2S. The transmission rate for methanol through our vinyl chloride-vinyl acetate copolymer is unknown; however, the rate for methanol through PVC%is 0.88 (g mil)/(h m2) and the rate for ethanol through CA27is 92 (g mil)/(h m2). For films 0.5-mil thick and 2 em2 in area these rates are equivalent to 2 X mol s-l through PVC and 2 X mol s-l through CA, both approximately fast enough to account for the rates of oxidation observed with each film and in the correct ratio within a factor of 2. We considered but discarded an alternative explanation, that 302permeated the films, was converted to IO2, and rediffused into solution to oxidize Bu2S. Neither the rates of diffusion of 302 into these films3snor the probable stability of '02 in the films39was sufficient to account for our measured rates of oxidation. Finally we consider the matter of sensitizer photodegradation. Recall that RB/CA films were far more susceptible to photobleaching than were RB/PVC films and that Bu2S photooxidation rates were also higher for RB/CA films. This indicates that sensitizer photodegradation and Bu2S photooxidation are coupled in some fashion and that O2 is involved in both processes. A mechanism consistent with this is given by reactions 3a, 3b, and 16
3S*+ ,02 3S*
S+.
-
+ ,02
-
-
lS + lo2
S+.
+ 02-.
colorless products
(3a) (3b)
(16)
According to this mechanism, excited-triplet sensitizer either by energy transfer (reaction 3a) to reacts with 302 give ground-state sensitizer plus '0, or by electron transfer (reaction 3b) to give sensitizer radical cation and superoxide. Subsequent reaction of s+.leads to photobleaching. Similar mechanisms for sensitizer photodegradation have been invoked elsewhere:@ and evidence for the occurrence of reaction 3b for RB30and has recently appeared. This interpretation also provides an intriguing rationale for the phenomenal stability of R ~ ( b p y to )~ photodegradation. For this sensitizer, reaction 3b leads to R~(bpy),~+. Ru(bpy)gB+is a strong reducing agent (1.2-V reduction potential in 1 M acid)32and could be expected to return to the 2+ oxidation state by reaction with an unspecified reductant rather than undergo subsequent
Photooxidation of Di-n-butyl Sulfide 813
Vol. 13, No. 4 , July--August 1980
Table VI1 Materials Used for Preparation of Polymer Films polymer cellulose acetate polystyrene poly( butyl methacrylate) poly(viny1 chloride)
manufacturer, type
solvent
Aldrich, acetyl content 39.8% Dow,Styron 666V du Pont, Elvacite 2044 Union Carbide, vinyl chloride-vinyl acetate copolymer type (VYHH)
90:10, methylene chloride-ethanol (95%) 75: 25: 25 toluene-methyl ethyl ketone-acetone cellosolve 50:45: 5 methylene chloride-acetone-ethanol (98%) (Rose Bengal) 50:40: 1 0 methvlene chloride-acetone-ethanol
reactions to yield colorless products.
Conclusions In the context of a system designed to mediate the photooxidation of organic substrates in solution, the incorporation of sensitizers i n t o solution-cast polymer films provides a ready alternative t o covalently b o u n d sensitizers. Using a cast film as s u p p o r t material permits the use of photosensitizers that could not otherwise be employed owing t o preparative difficulties. A good example of this is Ru(bpy),, a highly efficient sensitizer whose stability with respect t o photodegradation is far superior to that of typical organic dyes. While preparation of covalently b o u n d R u ( b p y ) s would entail difficult synthetic problems, cast films containing Ru(bpy), can be made with ease on as large a scale as desired. C a s t sensitizer films are more amenable t o investigations in which determination of quantum efficiencies is desired. There are distinct limitations t o t h e use of the cast sensitizer films, however. Some sensitizer/polymer combinations are suceptible t o leaching of the sensitizer i n t o solution. A related limitation is t h e reduction i n photooxidation quantum yield due to retarded t r a n s p o r t of solvent, dissolved oxygen, and/or s u b s t r a t e t h r o u g h the polymer film. Unfortunately, films that a r e highly permeable to solvent are also prone t o leaching of the sensitizer i n t o solution. S o m e compromise between quantum efficiency and resistance t o sensitizer leaching will therefore be necessary in t h e use of cast sensitizer films. T h i s is not to say that useful cast sensitizer films cannot be found. W e consider the Ru(bpy),/PVC material t o b e very useful i n converting Bu2S t o Bu2S0. We have extended the utility of the system t o photooxidation of neat BuzS and find that 90% conversion to BuzSO c a n b e obtained within 10 h b y using solar radiation and the Ru(bpy),/PVC film described in this report. Whether or not the cast sensitizer films will have a more general application to organic photooxidation awaits further investigation.
Experimental Section Materials. General Comments. Di-n-butyl sulfide (Bu,S) and di-n-butyl sulfone (Bu,SO,) from K & K Laboratories were shown by NMR and elemental analyses to be 97% pure. Water appeared to be the only major impurity in each compound, so they were used without further purification. Aldrich supplied 2,3-dimethyl-2-butene, which was used without further purification. Identities and relative amounts of contaminants in di-n-butyl sulfoxide (BuaO) from K & K Laboratories were determined by high-pressure liquid chromatography, using a refractive index detector, by NMR, and by UV-visible spectroscopy. The purity of the sample was taken into account when determining highpressure liquid chromatography response factors for analysis of reaction mixtures. All other materials were reagent grade and were used as supplied. Photosensitizers. The following photosensitizer dyes (color index number in parentheses) were purchased from Matheson Coleman and Bell and used without further purification: rose bengal (45440), saffranine 0 (502401, rhodamine B (45170), victoria
blue B (44045), 1-amino-4-hydroxyanthraquinone (60710), and malachite green (42000). Ru(bpy), was prepared from RuC1, and 2,2’-bipyridineby the procedure described by Sprintschnik et al.= Polymer Films. Sensitizer films were prepared from four different polymers: cellulose acetate, poly(buty1 methacrylate), polystyrene, and poly(viny1 chloride). All films were prepared by the solvent-casting technique; the polymer was dissolved in a solvent mixture and cast onto a glass plate with a “doctor’s blade”, a device that can be set to lay the solution a t any desired thickness. Typically, films were prepared as follows. A 10% solution of the polymer was made up, and sensitizer was added to this solution either by dissolving the sensitizer directly in the polymer solution or by mixing in a portion of a saturated solution cm thick by spreading of the dye. Films were made 2.5 X 6 g of solution onto a glass plate with the doctor’s blade set a t 2.0 X cm. The polymers, the manufacturers, and the solvents are listed in Table VII. Apparatus. Photochemical oxidations were performed in 10-mm i.d. X 13-cm 0.d. X 100-mm, screw-cap, borosilicate glass tubes (Pyrex 7740) mounted on a merry-go-round photochemical reactor (Ace Glass). The irradiation source was a Hanovia 450-W, medium-pressure Hg lamp contained in a borosilicate immersion well. The distance between the irradiation source and tubes was about 10 cm and the temperature was at the ambient operating temperature of the system, 30 f 2 “C. Two different sets of filters were used in the experiments. In solution-phase Bu,S oxidations with dissolved photosensitizers, 12-mm Corning 3-71 sharp-cut yellow filters were employed. These cut off below 440 nm and transmit more than 85% of the 546- and 578-nm Hg lamp emissions. For solution-phase experiments with sensitizer films, 4-mm Corning 4-102 band-pass filters were used. These filters transmit 13% of the 546-nm Hg line and cut off emissions above and below the 546-nm wavelength. The high-pressure liquid chromatography system (Waters Associates) consisted of a Model 660 solvent programmer, two Model 6000A pumps, and a U6K injector. Two different detectors were used in combination with this system: the Schoeffel Instrument Corp. GM 770 variable-wavelength UV detector and the Waters Associates R401 differential refractometer. Spectrophotometric measurements were performed on a Cary 15 UV-visible recording spectrophotometer and on a Spectronic 20 visible spectrophotometer. Cuvettes 1 cm in path length were used in all cases. Procedures. Actinometry. Absolute light intensities were determined with a potassium ferrioxalate After warming up the Hg lamp for 0.5 h, several 3-mL aliquots of aqueous 0.15 M K3Fe(C,04), were exposed to the filtered source for timed periods of 5-20 min. Then, 2-mL aliquots of these exposed solutions were diluted to volume with 0.19 M 1 , l O phenanthroline to form a colored complex with Fez+ions. Comparison of absorbance readings at 510 nm-with a Beer’s law curve for the phenanthroline complex yields values for Fe” formed during photolysis. The light intensity of the 546-nm Hg line may then be determined from eq 17, where ZOm is the intensity of the
546-nm Hg line incident on the photolysis cell (einstein L-’s-’), VI = 3 mL is the volume of K3Fe(C204)solution irradiated, V z = 2 mL is the volume of aliquot taken for analysis, V , = 25 mL is the final volume of phenanthroline solution to which V , is diluted, Aslo is the absorbance of (phenanthroline)iron(II)complex a t 510 nm, cbl0 = 1.04 X lo4 M-’ cm-’ is the molar extinction
814 Kenley, Kirshen, and Mill
Macromolecules
coefficient a t 510 nm for (phenanthroline)iron(II) complex as determined from the Beer’s law curve, 1 = 1 cm is the path length of the cuvette, I$F~M = 0.25 is the quantum yield for K3Fe(C204) photolysis a t 547 nm (see below), A,, = 0.015 is the absorbance at 546 nm of 0.15 M K3Fe(C204),and t is the photolysis time (9). Because the absorbance and photolysis quantum yield for 0.15 M K3Fe(C204)at 578 nm are negligible, it is necessary to calculate the intensity of the 578-nm Hg line from eq 18, where Z0578 is the
(18) intensity of the 578-nm Hg line incident on the photolysis cell (einstein L-’ s-l), E,,, is the rated% output of the Hanovia 450-W lamp at 578 nm, and E- is the rated output of the Hanovia 450-W lamp a t 546 nm. Since both the absorption coefficient and reported quantum yield for ferrioxalate photolysis at 546 nm ( I $ F p = 0.15) are small, we checked the system by simultaneously irradiating samples of ferrioxalate and a second chemical actinometer. The second actinometer was the RB-sensitized photooxidation of 2,3-dimethyl-2-butene to 2,3-dimethyl-2-(hydroperoxy)-2-butene (ROOH). This well-characterizedl photooxidation is similar to a chemical actinometer devised by Demas e t a1.,21except that we substituted RB (which has a higher absorption coefficient at 546 nm) for Ru(bpy), as the photosensitizer. The production of ROOH can be readily followed by iodimetric titration in acetic acid-chloroform solvent.37 Under the conditions employed (0.1 M 2,3-dimethyl-2-butene, 6.2 X lo4 M RB in oxygen-saturated methanol, 546-nm irradiation), all of the *02 produced by irradiation of RB is converted to ROOH, and the incident intensity (lo)can be calculated from eq 19, where (1 +d[ROOH]/dt
Io =
(1- 10-A%$o*
(19)
10-Am) = 0.61 is the fraction of light absorbed by the sensitizer and @02 = 0.8 is the quantum efficiency for the production of ‘02 from RB.’c*24An analogous expression (eq 20) can be written for
sensitizer solution. The amount of Bu2S added was chosen to give a solution approximately 5 X M in sulfide. Next, 2-mL aliquots of solution were transferred to photolysis tubes, and in each tube the volume over the solution was purged with 02.When prepared in this manner, each tube contained approximately 1 X mol of BuzS in solution, 2.4 X lo4 mol of O2 in solution, and 2.4 X lo4 mol of gaseous 02,over the solution. The solutions were irradiated for timed periods, and a syringe was used to withdraw accurately known volumes (10 pL) of solution for analysis. In some experiments, one tube was used for each time point; in others, a single tube with serum stopper was used for several time points. For photosensitizer films with Bu2S in methanol, solutions of M Bu2Swere prepared as above. Strips approximately 5 x of sensitizer films (1.1 X 3 cm) were cut and inserted into photolysis tubes. The tubes were purged with oxygen, and 2 mL of Bu2S solution was added. A series of tubes was irradiated, and individual tubes were withdrawn a t timed intervals. BuzS was analyzed by high-pressure liquid chromatography as above. In performing actinometry, it was necessary to correct for the fact that the illuminated surface area of the sensitizer films was less than the illuminated area of the photolysis tubes (including those used for actinometry) by a factor of 0.55. Moreover, during photodegradation of all sensitizer films, a certain amount of film blushing was observed. Since this blushing causes light scattering and high absorbance values, direct absorbance readings on these films were inaccurate. Therefore, we devised a procedure to determine these film absorbances indirectly. Exposed films were weighed and dissolved in 3 mL of methylene chloride, and the absorbance of the methylene chloride solution at 546 nm was measured. Unexposed films, the absorbances of which could be read directly, were handled in the same manner. After normalizing the absorbances on a weight basis, we could determine the exposed film absorbances by using eq 22, where is the absorbance (22) of exposed film after time t , A’,b is the absorbance of unexposed f i i read directly on the spectrophotometer, A? is the absorbance of a 3-mL solution of unexposed film, and A? is the absorbance of a 3-mL solution of exposed film after time t.
Acknowledgment. This work was performed under a the ferrioxalate actinometer, where -d[Fe3+]/dt is the rate of ferrioxalate reduction (determined as described above) and (110-Am) = 0.034 is the fraction of light absorbed by ferrioxalate. Combining eq 19 and 20 gives -d[Fe3+]/dt ‘Few
= +d[ROOH]/dt
For six samples of ferrioxalate irradiated for timed intervals of 600-3000 s, the average value of the rate of ferrioxalate reduction was -d[Fe3+]/dt = (1.22 & 0.05) X M s-’. For nine samples of 2,3-dimethyl-2-butene irradiated for ((t9) X lo3 s, the average rate of hydroperoxide production was +d[ROOH]/dt = (6.98 f 0.31) X lo4 M 8. Substituting these values into eq 21 yields a value of he^ = 0.25. This value of h e wwas used in conjunction with the ferrioxalate actinometer for all the experiments reported here. Chromatography. Concentrations of Bu2S and Bu2S0 were determined by high-pressure liquid chromatography. A 4-mm i.d. X 30 cm reverse phase WBondapak CIScolumn was used. The mobile phase was a CH3CN/H20 mixture (70:30). Bu2S and BuzSO were both determined with the UV detector a t 220 nm. Standard concentration vs. peak area curves were prepared and used to calculate Bu2S and Bu2S0 concentrations in reaction mixtures. Bu$302 exhibits a very weak molar extinction coefficient at 220 nm and could not be detected. Chromatographic peak areas were quantitated with a Spectra-Physics Minigrator digital integrator. Photooxidation. For photosensitizers and Bu2S in methanol, solutions of the sensitizers were prepared by using methanol saturated with O2 In each solution, the concentration of sensitizer was adjusted to give an absorbance of 0.50 a t 546 nm in a 1-cm path length cell. BuzS was accurately weighed into 25-mL volumetric flasks, and each flask was diluted to volume with a
c o n t r a c t with the Department of the Army (No. DAAKll-77-0055). We are indebted to Dr. Richard M. Laine for preparation of Ru(bpyI3, t o Mr. Dean Parkinson for assistance in preparation of polymer films, to Mrs.
Constance Gould for measurement of 02:Bu2S0stoichiometry, and to Dr. Joseph Epstein (US. Army Chemical Systems Laboratory) for many helpful suggestions.
References and Notes (a) C. S. Foote, Acc. Chem. Res., 1,104 (1968); (b) K. Gollnick and G. 0. Schenck in “1,4-Cycloaddition Reactions”, 3. Hammer, Ed., Academic Press, New York, 1967; (c) K. Gollnick, Adu. Photochem., 6 , l (1968); (d) D. R. Kearns, Chem. Rev., 71, 395 (1971). C. S. Foote in “Free Radicals in Biology”, Vol. 11, W. A. Pryor, Ed., Academic Press, New York, 1976. “Singlet Oxygen Reactions with Organic Compounds and Polymers”, B. Ranby and J. Rabec, Eds., Wiley, New York, 1978. E. C. Blossey, D. C. Neckers, A. C. Thayer, and A. P. Schaap, J . Am. Chem. SOC.,95, 3821 (1973). A. P. Schaap, A. L. Thayer, E. C. Blossey, and D. C. Neckers, J . Am. Chem. SOC.,97, 3741 (1975). D. C. Neckers, CHEMTECH, 8, 108 (1978). J. R. Williams, G. Orton, and L. R. Unger, Tetrahedron Lett., 4603 (1973). R. Nilsson and D. R. Kearns, Photochem. Photobiol., 19, 181 (1974). D. Brkic, P. Forzatti, I. Pasquon, and F. Trifiro, J . Mol. Catal., 3, 173 (1977/1978). J. Bourdon and B. Schnuriger, Photochem. Photobiol., 5,507 (1966). J. Bourdon, B. Schnuriger, and J. Bedu, Photochem. Photobiol., 8 , 361 (1968). S. Wolf, C. S. Foote, and J. Rebek, J . Am. Chem. SOC.,100, 7770 (1978).
Macromolecules 1980, 13, 815-820
(28) N. J. Turro. M. F. Chow, and M. A. Balustein, J . Am. Chem. SOC., 100, 7 i i i (1978). (29) K. J. Yautsev and L. I. Grossweiner, Photochem. Photobiol., 6, 721 (1967j. (30) V. S.Srinivasan, D. Podolsk, N. J. Westrick, and D. C. Neckers, J . Am. Chem. Soc., 100, 6513 (1978). (31) J. S. Winterle, D. S. Kliger, and G. S. Hammond, J . Am. Chem. SOC.,98,3719 (1976). (32) E. F. Lytle and D. M. Hercules, Photochem. Photobiol., 13,123 (1971). (33) G. Sprintschnik, H. W. Sprintschnik, P. Kirsch, and D. G. Whitten, J . Am. Chem. SOC.,99,4947 (1977). (34) J. G. Calvert and J. N. Pitts. “Photochemistry”, Wileu, New York, 1967, pp 783-6. (35) C. G. Hatchard and C. A. Parker, Proc. R. SOC.London, Ser. A , 235,518 (1956). (36) Hanovia Lamp Division, Canrad Precision Industries, Inc., Newark, NJ. (37) D. Swern in “Organic Peroxides”, Vol. 1, D. Swern, Ed., Wiley, New York, 1970, p 498. (38) Using known values for polymer permeability constants, we calculated that the rate of oxygen transport halfway into a PVC film (2.6 X mm thick; 2 mm2 area) at 30 OC and 2-torr O2pressure is 1.9 X W2mol s?, while under the same conditions for vinyl acetate-vinyl chloride copolymer it is 3.8 x lo-” mol s-l. (39) Bourdon et al.” and Turro et a1.28showed that ‘02can diffuse into polymer films only 500 A (5 x mm) before being quenched.
(13) M. W. Tembold and H. E. A. Kramer, J. SOC.Dyers Colour., 94,12 (1978). (14) V. B. Ivanov and V. Ya. Shlyapintokh, J. Polym. Sci., 16,899 (1978). (15) A. Zweie and W. A. Henderson. J . Polvm. Sci.. Polvm. Chem. Ed., 13,”717 (1975). (16) H.Kautsky, Trans. Faraday Soc., 35,216 (1939). (17) C. S. Foote and J. W. Peters, J. Am. Chem. Soc., 93,3795 (1 ,- 971) - . -,. (18) C. S.Foote and J. W. Peters. X X I I I Int. Conn. DD . - Pure A _ . ~. Chem., Spec. Lect. ZV, 129 (1971). (19) M. Cassanrande, G. Gennari, and G. Cauzzo, Gazz. Chin. Ital., 104,1251(197.4. (20) J. N. Demas, D. Diemente, and E. W. Harris, J . Am. Chem. SOC.,95,6864 (1973). (21) J. N. Demas, R. P. McBride, and E. W. Harris, J. Phys. Chem., 80,2248 (1976). J. N. Demas, E. W. Harris, and R. P. McBride, J . Am. Chem. Soc., 99,3547 (1977). For oxygen-saturated CH30H, typical values are k3(O2) = IO7 and k 4 = ~ O - ~ O ‘ S - ~ ? . . K. Gollnick. T. Franken. G. Schade.’ and G. Dorhofer. Ann. N.Y. Acad. Sci., 171, 89 ‘(1970). (25) M. Nemoto, H.Kokobun, and M. Koisumi, Bull. Chem. Soc. Jpn., 42, 1223 (1969). (26) C. E. Rogers in “Engineering Design for Plastics”, E. Baer, Ed., Reinhold, New York, 1964, pp 609ff. (27) J. Brandrup and E. H. Immergut, Eds., “Polymer Handbook”, Wiley-Interscience, New York, 1966. ~
815
~I
~I
Photochemistry of Polymers and Copolymers of Phenyl Vinyl Ketone and o-Tolyl Vinyl Ketone J. P. Bays, M. V. Encinas, and J. C. Scaiano*’ Radiation Laboratory,2 University of Notre Dame, Notre Dame, Indiana 46556. Received July 11, 1979
ABSTRACT: Energy migration plays an important role in the photochemistry of polymers containing phenyl vinyl ketone and o-tolyl vinyl ketone units. The average “residence time” of the triplet excitation in any given chromophore is in the neighborhood of 30 ps. Triplet lifetimes are estimated as 55 ns for PPVK and 2 ns for PTVK and are intermediate for all the copolymers. Biradical lifetimes are 65 ns for PPVK and 200 ns for PTVK, reflecting the longer lifetime of the biradicals derived from the o-methylbenzoyl group. Photoenolization of the 0-tolyl vinyl ketone moieties provides an energy sink, reducing the degree of photodegration. For example, a copolymer of phenyl vinyl ketone and 0-tolyl vinyl ketone containing 3% of the latter undergoes only 54% of the degradation undergone by PPVK.
The Norrish type I1 photofragmentation is an important process in polymer photo~hemistry.~,~ The reaction can take place in almost all polymers containing carbonyl groups, including some like polystyrene and polyethylene in which the carbonyl chromophores are incorporated as a result of oxidative processes which are common during polymerization, processing, and purification? Poly(pheny1 vinyl ketone), PPVK, is a good model for the Norrish type I1 process in macromolecules, and its photodegradation (reaction 1) has been the subject of several studie~.~,~-’l
T
j; jh” g
? +co Y co co I
I
Ph I
u
5.. -gH2 t
(1)
Ph
it will essentially behave as a mode of radiationless deactivation. In a recent preliminary communication12from this laboratory we have shown that photoenolizations in high polymers can be used to reduce, or prevent, photodegradation. For example, poly(o-tolyl vinyl ketone) is a photostable polymer; we have attributed this stability to the sequence of processes illustrated in reaction 2.
I
Ph
2
u
One way in which the energy can be channeled toward nondegradative processes is by providing an alternative reaction path for the excited carbonyl triplet, such that 0024-9297 /so/ 22 13-0815$01.oo/ 0
/
3 The behavior of copolymers of phenyl vinyl ketone and o-tolyl vinyl ketone, COPT, can be expected to be controlled by the relative rates of the various possible photoprocesses, as well as by the possibility of energy migra%
0 1980 American Chemical Society
